Page tree
  1. Load hits produced on the hitmaker step.
  2. Load alignment constants :
    1. Apply the planes' shifts obtained on the pre-alignment step (a part of the hitmaker step) to correct the hits' positions;
    2. Consecutively apply the planes' shifts obtained on the previous alignment step(s) to correct the hits' positions, i.e. on the alignment step n, apply n-1 sets of the planes' shifts obtained during previous n-1 alignment steps.
  3. Form track(s) out of the hits corrected on the step 2 :
    1. Make two groups of hits; each group consists of three hits corresponding to the planes of each telescope arm, i.e. three hits found on the upstream telescope arm (triplet) and three hits on the downstream arm (driplet);
    2. Combine triplet(s) and driplet(s) to track(s)
  4. Create the GBL trajectory(ies) out of the formed tracks
  5. Pass the created GBL trajectory(ies) to Millepede-II for a fine alignment of the (telescope) planes.
  6. Write the shift constants obtained by Millepede-II  per each plane of interest to an *.slcio file.

Steps 3-6 of the algorithm above are being processed by the EUTelMilleGBL.cc code.


Histograms description

A resulting *.root file after performing the align step has the following structure:

 

  • Proj_to_Plane0 : folder is contained by the \mathrm{x_{i}\,-\,x_{0}\,and\,y_{i}\,-\,y_{0}} distr., where i is the plane number;
  • Triplets : contains three sub-folders (Upstream_arm and Downstream_arm folders have the same structure)
    • Upstream_arm :
    • Downstream_arm :
    • StraightLine_tracks :
  • GBLtrajectories

 


Results

Comparison of the results depending on the number of align steps

Fig. 1.1. X residual distributions between the tel. plane hits and the GBL trajectory.

Fig. 1.2. The Mean value and Sigma distr. of the residual distr. Gaussian fit per tel. plane.

 

Fig 1.1, 1.2 show the X residual and their fits' parameters distr. per each tel. plane after different alignment step. Each alignment step was performed with the following cuts:

triCut = 60, driCut = 80, sixCut = 150, TelescopeResolution = 10.

Residual distr. "spread" doesn't decrease with additional alignment step(s), while the mean value is being fixed to ~0 after the first alignment step. The residual distr. spread rises up by a factor of ~10-15 due to the scattering at DUT (DUT material budged was estimated to be equal to 0.265, DUT inclination angle is 0 deg., in the units of rad length).

 

Fig. 2.1. X residual distributions between the tel. plane hits and the GBL trajectory.

Fig. 2.2. The Mean value and Sigma distr. of the residual distr. Gaussian fit per tel. plane.

Fig. 2.1, 2.2 show the X residual distr. and their fits' mean and sigma values after diff. align steps, the cuts which were used:

step 1: triCut = 60, driCut = 80, sixCut = 150, TelescopeResolution = 10; step 2,3 : triCut = 15, driCut = 20, sixCut = 20, TelescopeResolution = 4.

Step 2 gives a slight improvement for the residual distr. spread, while the step 3 makes the mean and spread worse for the downstream telescope arm's planes.

 


Presentations and scripts

tracking_alignment.pptx

tracking_alignment.vsdx

align.C

 

 

 

 

 

  • No labels